You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@tvm.apache.org by gi...@apache.org on 2023/01/11 05:02:42 UTC
[tvm] branch nightly updated (d00168ffbf -> 77b6f0eec3)
This is an automated email from the ASF dual-hosted git repository.
github-bot pushed a change to branch nightly
in repository https://gitbox.apache.org/repos/asf/tvm.git
from d00168ffbf [CI] Fix MLF input and output name map (#13740)
add 8d53c0aa8a [TE][PrimFunc] Fix create primfunc from te extern with explicit buffer load (#13729)
add 1265eb93e7 Add Name Transforms for Rust style (#13706)
add db920ddcde [COLLAGE] Add more customization to support more targets (#13450)
add 3b5baf48d2 [Docker update] Update ci_cpu tag to the latest from tlcpackstaging (#13748)
add 746fcaaaa3 [CMSIS-NN] Support CMSIS NN from new GitHub location (#13656)
add 4cb75b97cd [microNPU] Add a legalization test for TFLite PAD (#13750)
add 68c917d9b4 [Arith] Use ConstIntBound to remove negative numerator when lowering (#13724)
add 92da138bcb [Profiler] Allow user to flush L2 cache in `time_evalutor` function for profiling CUDA kernels (#13726)
add d979949993 [CMake][OpenCL] Remove warning for OpenCL wrapper (#13683)
add bf4fedd06e [AOT]Aot module post-test error workaround (#13685)
add c2bc1ec95e [microTVM] tuning on micro targets with meta-schedule (#13514)
add 77b6f0eec3 [Tensorize][TOPI] Add AMX Tensorizing for int8 batch matmul (#13745)
No new revisions were added by this update.
Summary of changes:
3rdparty/nvbench/l2_cache_flush.h | 74 +++++
LICENSE | 2 +-
apps/microtvm/cmsisnn/Makefile | 9 +-
apps/microtvm/cmsisnn/README.md | 1 +
apps/microtvm/ethosu/Makefile | 9 +-
apps/microtvm/ethosu/README.md | 1 +
.../template_project/CMakeLists.txt.template | 26 +-
apps/microtvm/zephyr_cmsisnn/CMakeLists.txt | 44 ++-
apps/microtvm/zephyr_cmsisnn/README.md | 5 +
ci/jenkins/docker-images.ini | 2 +-
cmake/modules/OpenCL.cmake | 2 +-
docker/install/ubuntu_install_cmsis.sh | 4 +
.../install/ubuntu_install_ethosu_driver_stack.sh | 3 +
include/tvm/meta_schedule/mutator.h | 2 +
include/tvm/meta_schedule/postproc.h | 2 +
include/tvm/meta_schedule/schedule_rule.h | 2 +
...CENSE.dlpack.txt => LICENSE.l2_cache_flush.txt} | 61 ++--
python/tvm/contrib/hexagon/session.py | 5 +-
.../micro/meta_schedule/local_builder_micro.py | 84 +++++
.../micro/meta_schedule/rpc_runner_micro.py | 233 ++++++++++++++
python/tvm/meta_schedule/relay_integration.py | 30 +-
python/tvm/micro/project_api/server.py | 2 +-
python/tvm/relay/backend/executor.py | 13 +
python/tvm/relay/collage/__init__.py | 1 +
python/tvm/relay/collage/collage.py | 8 +
python/tvm/relay/op/contrib/clml.py | 20 ++
python/tvm/relay/op/strategy/x86.py | 10 +-
python/tvm/rpc/server.py | 30 +-
python/tvm/rpc/tracker.py | 25 +-
python/tvm/topi/x86/batch_matmul.py | 53 ++-
python/tvm/topi/x86/dense.py | 21 +-
src/meta_schedule/mutator/mutator.cc | 8 +
src/meta_schedule/postproc/postproc.cc | 8 +
src/meta_schedule/schedule_rule/schedule_rule.cc | 27 ++
.../space_generator/space_generator.cc | 7 +
src/meta_schedule/utils.h | 2 +
src/relay/backend/name_transforms.cc | 51 ++-
src/relay/backend/name_transforms.h | 24 ++
src/relay/collage/collage_partitioner.cc | 2 +-
...{cost_estimator.cc => custom_cost_estimator.cc} | 31 +-
src/relay/collage/custom_cost_estimator.h | 67 ++++
src/relay/collage/gather_partition_specs.cc | 35 +-
src/relay/collage/utils.cc | 13 +
src/relay/collage/utils.h | 6 +
.../{name_transforms.cc => cuda/l2_cache_flush.cc} | 26 +-
src/target/source/codegen_c_host.cc | 1 +
src/te/operation/create_primfunc.cc | 43 ++-
src/tir/ir/specialize.cc | 6 +-
src/tir/transforms/lower_intrin.cc | 136 ++++++--
tests/cpp/name_transforms_test.cc | 50 ++-
tests/micro/zephyr/test_ms_tuning.py | 171 ++++++++++
tests/micro/zephyr/test_zephyr.py | 3 +-
.../test_clml/test_adreno_collage_targets.py | 354 +++++++++++++++++++++
tests/python/contrib/test_ethosu/test_legalize.py | 100 ++++++
tests/python/relay/aot/corstone300.mk | 9 +-
.../relay/collage/demo_collage_partitioner.py | 6 +
tests/python/relay/test_op_level10.py | 55 ++++
.../python/unittest/test_evaluator_with_preproc.py | 60 ++++
tests/python/unittest/test_micro_ms_tuning.py | 126 ++++++++
.../python/unittest/test_target_codegen_vulkan.py | 42 +++
tests/python/unittest/test_te_create_primfunc.py | 30 ++
61 files changed, 2102 insertions(+), 181 deletions(-)
create mode 100644 3rdparty/nvbench/l2_cache_flush.h
copy licenses/{LICENSE.dlpack.txt => LICENSE.l2_cache_flush.txt} (82%)
create mode 100644 python/tvm/contrib/micro/meta_schedule/local_builder_micro.py
create mode 100644 python/tvm/contrib/micro/meta_schedule/rpc_runner_micro.py
copy src/relay/collage/{cost_estimator.cc => custom_cost_estimator.cc} (61%)
create mode 100644 src/relay/collage/custom_cost_estimator.h
copy src/runtime/{name_transforms.cc => cuda/l2_cache_flush.cc} (63%)
create mode 100644 tests/micro/zephyr/test_ms_tuning.py
create mode 100644 tests/python/contrib/test_clml/test_adreno_collage_targets.py
create mode 100644 tests/python/unittest/test_evaluator_with_preproc.py
create mode 100644 tests/python/unittest/test_micro_ms_tuning.py